Recent developments concerning the rapid single-flux-quantum (RSFQ) circuit family are reviewed. Elementary cells in this circuit family can generate, pass, memorize, and reproduce picosecond voltage pulses with a nominally quantized area corresponding to transfer of a single magnetic flux quantum across a Josephson junction. Functionally, each cell can be viewed as a combination of a logic gate and an output latch (register) controlled by clock pulses, which are physically similar to the signal pulses. Hand-shaking style of local exchange by the clock pulses enables one to increase complexity of the LSI RSFQ systems without loss of operating speed. The simplest components of the RSFQ circuitry have been experimentally tested at clock frequencies exceeding 100 GHz, and an increase of the speed beyond 300 GHz is expected as a result of using an up-to-date fabrication technology. This review includes a discussion of possible future developments and applications of this novel, ultrafast digital technology. >

RSFQ logic/memory family: a new Josephson-junction technology for sub-terahertz-clock-frequency digital systems

A pipeline video decoder and decompression system handles a plurality of separately encoded bit streams arranged as a single serial bit stream of digital bits and having separately encoded pairs of control codes and corresponding data carried in the serial bit stream. The pipeline system employs a plurality of interconnected stages to decode and decompress the single bit stream, including a start code detector. When in a search mode, the start code detector searches for a specific start code corresponding to one of multiple compression standards. The start code detector responding to the single serial bit stream generates control tokens and data tokens. A respective one of the tokens includes a plurality of data words. Each data word has an extension bit which indicates a presence of additional words therein. The data words are thereby unlimited in number. A token decode circuit positioned in certain of the stages recognizes certain of the tokens as control tokens pertinent to that stage and passes unrecognized control tokens to a succeeding stage. A reconfigurable decode and parser processing means positioned in certain of the stages is responsive to a recognized control token and reconfigures a particular stage to handle an identified data token. Methods relating to the decoder and decompression system include processing steps relating thereto.

Multistandard video decoder and decompression system for processing encoded bit streams including a video formatter and methods relating thereto

Current microprocessors employ a global timing reference to synchronize data transfer. A synchronous system must know the maximum time needed to compute a function, but a circuit usually finishes computation earlier than the worst-case delay. The system nevertheless waits for the maximum time bound to guarantee a correct result. As a first step in achieving variable pipeline delays based on data values, approximation circuits can increase clock frequency by reducing the number of cycles a function requires. Instead of implementing the complete logic function, a simplified circuit mimics it using rough calculations to predict results. The results are correct most of the time, and simulations show improvements in overall performance in spite of the overhead needed to recover from mistakes.

Speeding up processing with approximation circuits

Programmability is an important requirement for portable computing and communication devices that must be flexible enough to accommodate a variety of multimedia services and communication capabilities. However, compared to dedicated, application-specific solutions, programmable devices often incur significant performance and power penalties. We present a hybrid architecture template that can be used to implement ultra-low-power programmable processors for signal processing applications.

Ultra-low-power domain-specific multimedia processors

A pipeline video decoder and decompression system handles a plurality of separately encoded bit streams arranged as a single serial bit stream of digital bits and having separately encoded pairs of control codes and corresponding data carried in the serial bit stream. The pipeline system employs a plurality of interconnected stages to decode and decompress the single bit stream, including a start code detector. When in a search mode, the start code detector searches for a specific start code corresponding to one of multiple compression standards. The start code detector responding to the single serial bit stream generates control tokens and data tokens. A respective one of the tokens includes a plurality of data words. Each data word has an extension bit which indicates a presence of additional words therein. The data words are thereby unlimited in number. A token decode circuit positioned in certain of the stages recognizes certain of the tokens as control tokens pertinent to that stage and passes unrecognized control tokens to a succeeding stage. A reconfigurable decode and parser processing means positioned in certain of the stages is responsive to a recognized control token and reconfigures a particular stage to handle an identified data token. Methods relating to the decoder and decompression system include processing steps relating thereto

Data pipeline system and data encoding method

An elastic pipeline mechanism that is especially suitable for data-driven processors is described. With the elastic pipeline scheme, a large processing rate and a smooth data stream in the pipeline are realized at the same time. Two types of self-timed circuits, which are used for data-transfer control circuits in the elastic pipeline, are proposed. Using different types of transfer control circuits, two loop-shaped elastic pipeline mechanisms have been implemented on test chips and are compared with each other. One of these chips demonstrated that the data throughput in the pipeline was 55 megawords per second and that the critical path within a pipeline stage corresponded to 16 inverter delays. This indicates the possibility of high-performance data-driven processors. >

An elastic pipeline mechanism by self-timed circuits

A microprocessor which can execute a test and set instruction for an exclusive control by combination of a few simple instructions, and data flow microprocessor which realizes high operation performance mainly in vector operation by reading out of data to be operated, writing in operation result and executing memory access in short time period and in parallel, and whose running efficiency of program is high in multi-processor construction.

Microprocessor and data flow microprocessor having vector operation function

A data driven processing system includes a mechanism for generating a data pair from a sequential input data stream by matching identifier fields. The pairing mechanism comprises a hash memory in which input data words to be paired are stored by using hashed addresses. If a hash collision occurs, the data word which caused the hash collision is transmitted to a counter-directional data loop which is used to generate a data pair. If an input data word is not paired after one pass through the data loop it is returned to the hash memory for another pairing operation. Use of both the hash memory and the counter-directional data loop reduces the required hash memory size and increases processing efficiency.

Data driven processor with data pairing apparatus combining a hash memory with counter directional data loops

A data-driven processor which has a packet assembling unit to add a tag information to the sequentially inputted data when the input data has no tag information, such as destination address or the like, thereby enabling the data to be inputted without using an external circuit, such as a host processor, and improving the data input rate and which also has a packet outputting and rearranging unit for rearranging an output packet stream in a predetermined order to thereby output the data information only, so that it is possible that the data is outputted without any external circuit, such as a host processor, the output rate is improved and the data output is executed in a predetermined order.

Data-driven processor having an internal tag-generating system for generating a distinct tagged information and assembling with un-tagged information of an input/output data packet

PURPOSE:To improve a data input/output rate and to decrease the number of input/output pins by providing a packet assembly part which adds tag information to input data having no tag information and assembles a packet and an output control part which outputs a packet processed by an arithmetic processing part. CONSTITUTION:The packet assembly part 50 adds ID numbers starting with 0 to data which are inputted in order and have no tag in order to generate tag information and a packet consisting of the tags and data is outputted to a link table part 2. Then a cyclic pipeline process consisting of a function table part 3, a queue memory part 4, a queue part 5, an arithmetic part 6, and a link table part 2 is performed and an output control part 55 rearranges arrival packets in ID number order through an output queue part 7. Only data part which have no tag parts are outputted in specific order specified according to the ID numbers. Consequently, an increase in the number of pin parts due to the input and output of tag parts and a decrease in input/output rate are precluded.

F. Asai

Papers

An elastic pipeline mechanism by self-timed circuits

Microprocessor and data flow microprocessor having vector operation function

Data driven processor with data pairing apparatus combining a hash memory with counter directional data loops

Data-driven processor having an internal tag-generating system for generating a distinct tagged information and assembling with un-tagged information of an input/output data packet

Data driving type computer